21-Feb-92 Tom Horton:ESCP10 Highlight color info from Rumph and Laman
Date: 21-Feb-92 13:37:49 PST Subject: Highlight color info from Rumph and Laman Message-ID: Originator: "Tom Horton:ESCP10:xerox", UniqueString: "21-Feb-92 13:37:49 PST" To: plass:parc Cc: Tom Horton From: Tom Horton:ESCP10:xerox
Located on [USC:OSBU South:Xerox]<Public USC>thorton> is ImagerPrintColorImpl.mesa which provides a bug fix for the highlight context. I believe it is a change from SetGray to a SetFraction call in the constantColorOp. I will also place there a folder from Bob Laman with the default entire setcolorscreen and the 3 tables.
All of these things are located on blintz except the setcolorscreen information if that it easier to access. The 3 tables are /PCedarDev2/wasabi/tom/siena/ and the ImagerPrintColorImpl.mesa and ImagerPrintColorImpl.dif are on /casa/horton/
As far as the data files. There is one each for each of the possible highlight toner colors that might be on the printer. They are mappings from RGB values to highlight values. The incoming RGB values are 6 bits Red, 6 Green and 6 Blue. The resulting values are 8 bit Highlight and 8 bit Black. Each file therefore is 64 cubed x 2 or 524288 bytes. The RGB values represent intensity as defined in the Redbook with 0 0 0 equal to black and 1 1 1 to white, yet the data is specified in the amount of ink. Therefore, a decimal RGB value of 20 20 20 would find a black-highlight word (but in reality it looks to be a highlight-black word--I am checking on this with Bob Laman) at file location 166440 decimal.
The Entire setcolorscreen default spot function among other things is 50 halftone cells per inch @ 45 degrees.
Tom
- - - - - - - - - - - - - - - From: John D. Halbur:ES AE:Xerox Sender: John D. Halbur:ES AE:Xerox Subject: Siena Profile Date: 24-October-90 (Wednesday) 13:07:59 PDT To: ryard, mmahmud Copies: halbur, alfvin In reply to: John D. Halbur:ES AE:Xerox Note: Rick and Max,
It looks like the Font Solution printer profile for Siena will use no thickening and a minimum stroke width of 1.5. This means that a single pixel wide line on Siena will show up much better than on the other 4000 printers and that we will have to address AR 550, and any other fudging we do to bump up stroke widths, for the 1.4 release. For a halftone file, we should probably try a neutral NTX compatible dot.
/John
19-Mar-92 Tom Horton:ESCP10 Re: Highlight color info from Rumph and Laman
Date: 19-Mar-92 15:15:32 PST Subject: Re: Highlight color info from Rumph and Laman In-Reply-To: "Michael Plass:PARC's message of 19 Mar 92 14:12:34 PST (Thursday)" Message-ID: Originator: "Tom Horton:ESCP10:xerox", UniqueString: "19-Mar-92 15:15:32 PST" To: Michael Plass:PARC Cc: Tom Horton, plass:PARC From: Tom Horton:ESCP10:Xerox
Michael, Following is the Holladay matrix for emulating the ENTIRE halftone dot. The highlight dot was derived from the black dot, I havenıt yet verified that ENTIRE does the same thing. I think that ENTIRE processes in Landscape for SIENA, but the Portrait mode is included in case you need it. -- Chuck The ENTIRE Postscript dot for Landscape processing, black [ [179 84 100 68 44 203 171 131] [147 124 28 20 108 187 251 235] [219 52 4 12 36 155 243 227] [211 60 92 116 76 139 163 195] ] 4 255 black brick
The express pack with the second set of samples arrived this afternoon; I will see how they print with our version of the table driven mapping, probably sometime next week.
I learned from Chuck's phone message that you provided him with an RES file of Entire output; if you (or Chuck) could provide me with that, together with the postscript that made it, I might have a chance of knowing whether I've got it right. Comparing the output of out 4850 with that of your 4850 is not a very robust way of closing the loop.
Although the highlight mapping tables and Entire-compatible dots made it into the PARC C02 delivery to XSIS, I am not at all confident that they are right. Any necessary corrections will follow in an update.
Action item: Provide me with postscript + RES file of Entire output.
Michael, You were right, the highlight color dot growth is different. Iıll send it soon, itıs more difficult to deduce because some of the levels never get produced by ENTIRE.
The PostScript spot function has the highlight color as theŞ invertŞ of the black, but I suppose that as the two functions go through the 45 degree rotation the bits get clobbered. Unfortunately, since it is not really the inverse, there are a lot of overlapping pixels. SIENA makes the overlapping pixels disappear for the highlight color.
This dot is a real disaster. The dot-growth is not well-clustered but somewhat fractured. The isolated pixels in the dot growth are very susceptible to machine speed and xerographic problems, and are guaranteed to be very noisy. The highlight dot is not centered in the holes of the black dot. Some of the highlight levels will be smaller when they are printed together with black.
I understand the need to be compatible with ENTIRE, but I really think that this emulation is going too far. Perhaps the mapping that you are working on will be enough to insure compatibility. If we use a nice quiet well-clustered Holliday dot with the same frequency and number of levels, the area coverage for each level should be close enough to the ENTIRE dot, especially since the ENTIRE dot is niosy anyway.
I will make a dot that is as close as possible to the ENTIRE dot but without the problems. How about it? -- Chuck
15-Jun-92 Chuck Hains:XSIS Entire IMGıs and AISıs
Date: 15-Jun-92 16:16:18 Sender: Chuck Hains:XSIS:Xerox Subject: Entire IMGıs and AISıs Message-ID: Originator: "Chuck Hains:XSIS:Xerox", UniqueString: "15-Jun-92 16:16:18" To: Michael Plass:PARC Cc: Hains From: Chuck Hains:XSIS:XEROX
Michael, The two Imgıs that Bob Laman sent me are on the xns server /OrangeGrove-2:XSIS:Xerox/General Public/for Michael Plass/black.img, ...red.img, and as ais files also.
Date: 16-Jun-92 18:04:04 Sender: Chuck Hains:XSIS:Xerox Subject: ENTIRE halftone emulation Message-ID: Originator: "Chuck Hains:XSIS:Xerox", UniqueString: "16-Jun-92 18:04:04" To: Michael Plass:PARC Cc: RLaman:LAX1B, ASaran:LAX1B, PrintersRUs:PARC, Tom Horton:ESCP10, Wei-Jhy Chern:ESCP10, Steven S. Kang:ESCP10, Richard K. Yardumian:ESCP10, MMahmud:ESCP10, WChern:ESCP10 Reply-To: Chuck Hains From: Chuck Hains:XSIS:Xerox
Michael, Following is an updated version of the Holladay matrix for emulating the ENTIRE halftone dot. The highlight dot is different from the last version.
Both the black and the highlight dots were derived from actual IMG files that were produced by the ENTIRE implementation.
Also included is a set of dots that I think we should use instead of the ENTIRE dot, but with the same color mapping that you have implemented. -- Chuck The ENTIRE Postscript dot for Landscape processing, Black & Highlight [ [179 84 100 68 44 203 171 131] [147 124 28 20 108 187 251 235] [219 52 4 12 36 155 243 227] [211 60 92 116 76 139 163 195] ] 4 255 black brick